Chris Wolf, vice president, VMware AI Labs spoke to Dataquest at the VMware Explore 2023 Singapore. He said that it’s essential to consider AI platforms that offer flexibility, given the rapidly evolving landscape of AI technology. “Instead of succumbing to the pressure of committing to a single software stack hastily, prioritize having multiple options at your disposal. Furthermore, keep a close eye on the burgeoning trend of smaller AI models and the traction they are gaining. We believe that these compact models, finely tuned for specific business processes, will play a significant role in the future and become increasingly prevalent. This is an area that warrants close attention as it continues to evolve,” stated Chris.
[Excerpts from the Interview]
DQ: How is generative AI making an impact in industries like healthcare, finance, and manufacturing?
Chris Wolf: It’s quite remarkable. I’m genuinely enthusiastic about the potential of healthcare and Generative AI. It plays a pivotal role in exposing healthcare professionals to an unprecedented volume of data, something they couldn’t access in several lifetimes of medical practice. This technology enables healthcare practitioners to arrive at diagnoses much faster, ultimately enhancing patient safety and early detection. To illustrate, one of the standout statistics is related to breast cancer screening, where AI-assisted screenings have demonstrated a staggering 30% increase in accuracy. That’s a substantial improvement. In the realm of manufacturing, we’re witnessing notable applications in worker safety and early system failure detection for proactive maintenance. Banking, with its extensive history in data analytics, is becoming increasingly adept at fraud detection, which, in turn, benefits areas like document and loan processing. AI is permeating all facets of life, even academia, where my older children are encouraged by their professors to utilize generative AI tools for their assignments, recognizing its relevance as a professional skill they’ll carry forward. It’s essentially a tool they’ll rely on in their future careers, and it’s advantageous for them to acquire proficiency during their academic years.
DQ: In which industries do you see the most potential for Private AI or generative AI adoption in the near future?
Chris Wolf: Certainly, we’re actively engaged across a wide spectrum of industries. In the realm of code development, the reach of AI extends across virtually any vertical, especially within the tech sector, but we’re also witnessing its impact in manufacturing and various other sectors, which is quite promising. Additionally, healthcare is an area where we’re already seeing remarkable early successes as far as AI is concerned, and it’s truly exhilarating. Doctors, for instance, are eager to enhance their ability to diagnose patients more effectively. Meanwhile, in the retail sector, one of my personal favorites, computer vision is being employed to identify when a customer is facing difficulties in locating their desired products. This enables human assistance to be precisely directed to those customers in need, ultimately boosting sales—a win-win situation all around.
DQ: There is a lot of excitement on Private AI. However, What are the primary challenges that different industries face when adopting AI technologies, and how do you plan to help them in overcoming these hurdles?
Chris Wolf: Certainly, the challenges in the AI landscape are multifaceted, including a shortage of the requisite skill set and a general lack of understanding regarding the feasibility of AI use cases. Some commonly held assumptions, like the belief that on-premises AI deployments necessitate thousands of GPUs, need to be dispelled. In fact, we’ve successfully executed AI use cases with just two GPUs serving hundreds of users, yielding outstanding results. Overcoming these barriers entails a significant educational effort, and VMware is actively contributing by sharing our insights and knowledge garnered from running internal AI use cases on a global scale. Additionally, there’s the issue of organizations hesitating due to the rapid pace of AI advancement, fearing they might make the wrong choice. To address this concern, VMware has taken a proactive approach by offering a private AI infrastructure, ensuring future-proofing through software updates rather than causing fear of making incorrect decisions. There are alternative approaches available to navigate this challenge effectively.
DQ: Aren’t data security laws counterproductive to Generative AI?
Chris Wolf: I don’t view regulations as inherently counterproductive. While they may pose challenges for certain business models, VMware’s private AI architecture is designed to align AI models and computing resources with the specific data requirements, ensuring flexibility. This approach enables us to cater to customers and industries on their own terms. Rather than opposing regulation, I believe it serves a valuable purpose. Regulations are established with the intention of safeguarding people’s privacy and the confidentiality of their data, which is a commendable objective.
DQ: How do you see the future of generative AI adoption evolving, and what are the potential long-term impacts on the workforce, economic landscape, and society at large?
Chris Wolf: Shifts in industries are constant occurrences, and that’s simply the nature of progress. Take the energy sector, for example; today, there are fewer coal miners, but there are tens of thousands of jobs in solar production, wind energy, and other alternative energy sources. Throughout history, technology has consistently brought about changes in employment patterns, and we are currently experiencing another transformation. People will adapt and harness these tools to become more productive in ways they might not have envisioned before. Furthermore, considering the substantial time professionals spend in offices, if AI can help us accomplish tasks more efficiently and improve our work-life balance, I’m wholeheartedly in favor of it. It’s a positive development.
DQ: As industries adopt generative AI there is growing demand. How are you planning to meet this demand?
Chris Wolf: Indeed, it’s a significant challenge. The hardware industry is presently racing to align itself with the soaring demand. Companies like NVIDIA are grappling with substantial backlogs in shipping their GPUs, while Intel and AMD are intensifying the production of CPUs and their own GPU lines to meet the growing needs of the market. As production scales up, it will eventually catch up with demand. From VMware’s standpoint, our emphasis is on providing options. We advocate that making an investment in a virtual infrastructure equips you with the flexibility to utilize any hardware you require both now and in the future. This strategic virtual infrastructure investment today will expand your choices and ensure adaptability to changing technology landscapes.
DQ: Finally, what advice do you have for businesses and professionals looking to embrace generative AI in their respective industries? What steps should they take to ensure a smooth and responsible transition?
Chris Wolf: Absolutely, there’s a wealth of possibilities in the realm of AI adoption. My recommendation is to start by honing in on a quick win. Select one use case that can provide tangible benefits and deploy it, learning from the experience. It’s not necessary to apply AI to every facet of your operations. Similar to the initial days of blockchain, where there was an abundance of hype and an inclination to apply it indiscriminately, a more discerning approach is required for AI. Focus areas such as code development, content development for sales and marketing, and the implementation of chatbots for enhanced customer service and support are excellent starting points. Additionally, ensuring consistency in sales contracts and legal agreements can be facilitated through Generative AI. These are practical targets for organizations. Secondly, at VMware, we’ve established an AI governance council at the beginning of the year. This council comprises not only technology experts but also key figures such as our head of legal, chief information security officer, and representatives from various business lines. Together, they make decisions regarding the establishment of appropriate guardrails for AI usage within the company. This approach ensures that company employees are well-informed about AI guidelines and provides a mechanism for addressing exceptions that may not be covered by existing guidelines. This cross-company governance is vital for ensuring the safe and responsible use of AI, involving crucial stakeholders such as the Chief Information Security Officer (CISO) and legal experts, even if their role involves saying “no” at times, as it ultimately safeguards the company’s interests.